Distribution based on PAK

[Blue Like Jazz]

Number correct distribution from 102 PAK submissions versus SOA released answer key.

Sorry, the column labels suck. You get the idea.

# n N? F S
1 0 0 0% 100%
2 0 0 0% 100%
3 0 0 0% 100%
4 0 0 0% 100%
5 0 0 0% 100%
6 0 0 0% 100%
7 0 0 0% 100%
8 0 0 0% 100%
9 0 0 0% 100%
10 0 0 0% 100%
11 0 0 0% 100%
12 2 2 2% 98%
13 0 2 2% 98%
14 3 5 5% 95%
15 1 6 6% 94%
16 2 8 8% 92%
17 3 11 11% 89%
18 2 13 13% 87%
19 3 16 16% 84%
20 3 19 19% 81%
21 4 23 23% 77%
22 5 28 27% 73%
23 3 31 30% 70%
24 10 41 40% 60%
25 7 48 47% 53%
26 5 53 52% 48%
27 12 65 64% 36%
28 8 73 72% 28%
29 7 80 78% 22%
30 9 89 87% 13%
31 6 95 93% 7%
32 1 96 94% 6%
33 1 97 95% 5%
34 1 98 96% 4%
35 2 100 98% 2%
36 1 101 99% 1%
37 1 102 100% 0%
38 0 102 100% 0%
39 0 102 100% 0%
40 0 102 100% 0%

[TiderInsider]

Isn't the 2nd to last column percentile? Anyway, I think it will be interesting to see how many PAK submitters pass vs the actual pass percentage.

[Wannabe Actuary]

if 24 is passing then 71/102 PAK submitters passed. But keep in mind, some people, like myself, that had trouble with the exam, spent time on the problems, not taking our answers out. I would bring that percent down. So the results are biased anyway...i.e. if you do well, you have time to store your answers.

[Blue Like Jazz]

Originally Posted by TiderInsider

Isn't the 2nd to last column percentile?

You are correct. My deepest apologies. I will fix.

[BallaActuary]

Plus, if you answer only 15 questions or less, there's a good chance you didn't pass and aren't going to waste your time sending them to a PAK. I'd believe that people who submit their answers always have a greater pass ratio than the population of exam takers - which is good because it means the PAK will be fairly accurate.

[Blue Like Jazz]

Sorry for the crappy formatting.

November 2005 vs May 2005

# n N F S # n N F S
1 0 0 0% 100% 1 0 0 0% 100%
2 0 0 0% 100% 2 0 0 0% 100%
3 0 0 0% 100% 3 0 0 0% 100%
4 0 0 0% 100% 4 0 0 0% 100%
5 0 0 0% 100% 5 0 0 0% 100%
6 0 0 0% 100% 6 0 0 0% 100%
7 0 0 0% 100% 7 0 0 0% 100%
8 0 0 0% 100% 8 0 0 0% 100%
9 0 0 0% 100% 9 0 0 0% 100%
10 0 0 0% 100% 10 0 0 0% 100%
11 0 0 0% 100% 11 0 0 0% 100%
12 2 2 2% 98% 12 0 0 0% 100%
13 0 2 2% 98% 13 0 0 0% 100%
14 3 5 5% 95% 14 0 0 0% 100%
15 1 6 6% 94% 15 0 0 0% 100%
16 2 8 8% 92% 16 2 2 4% 96%
17 3 11 11% 89% 17 0 2 4% 96%
18 2 13 13% 87% 18 0 2 4% 96%
19 3 16 16% 84% 19 0 2 4% 96%
20 3 19 19% 81% 20 1 3 6% 94%
21 4 23 23% 77% 21 4 7 15% 85%
22 5 28 27% 73% 22 1 8 17% 83%
23 3 31 30% 70% 23 2 10 21% 79%
24 10 41 40% 60% 24 4 14 29% 71%
25 7 48 47% 53% 25 5 19 40% 60%
26 5 53 52% 48% 26 3 22 46% 54%
27 12 65 64% 36% 27 3 25 52% 48%
28 8 73 72% 28% 28 3 28 58% 42%
29 7 80 78% 22% 29 7 35 73% 27%
30 9 89 87% 13% 30 1 36 75% 25%
31 6 95 93% 7% 31 2 38 79% 21%
32 1 96 94% 6% 32 1 39 81% 19%
33 1 97 95% 5% 33 3 42 88% 13%
34 1 98 96% 4% 34 2 44 92% 8%
35 2 100 98% 2% 35 1 45 94% 6%
36 1 101 99% 1% 36 2 47 98% 2%
37 1 102 100% 0% 37 1 48 100% 0%
38 0 102 100% 0% 38 0 48 100% 0%
39 0 102 100% 0% 39 0 48 100% 0%
40 0 102 100% 0% 40 0 48 100% 0%

[TiderInsider]

Originally Posted by BallaActuary

Plus, if you answer only 15 questions or less, there's a good chance you didn't pass and aren't going to waste your time sending them to a PAK. I'd believe that people who submit their answers always have a greater pass ratio than the population of exam takers - which is good because it means the PAK will be fairly accurate.

When I said that I thought it would be interesting to see the difference between actual pass % and PAK pass %, I didn't mean that I had no clue which one would be higher. I'd like to see how much higher.

[TiderInsider]

Originally Posted by justme

November 2005 vs May 2005

Is this based off of May 2005's PAK responders? Looks like the sample size is less than half of this year.

[Blue Like Jazz]

yes.

[Blue Like Jazz]

but it was still pretty accurate.